Auto-Components Plant
₹42 lakh/yr verified savings · rejections down 44%
The starting point
The plant machines transmission components for two large OEMs. On paper it ran a lean program: monthly 5S audits, a suggestion box, a quality board in the canteen. In practice, the audits were clipboard forms filled in at a desk, the box collected dust, and the same oil leak under CNC line 2 was 'found' at every audit for a year — noted, never fixed.
The pressure arrived from outside: a key customer moved to stricter incoming inspection and rejections started creeping past target. Every rejected lot triggered a firefight — sort, rework, apologize — but the causes were never traced, so the same defects kept coming back.
What changed, step by step
LeanOS runs one connected loop — audit → defect → fix → root cause → kaizen → standard → recognition. Here is what each step did at this plant.
Clipboards out, phones out
Supervisors mapped the plant into 14 zones in an afternoon. Instead of a monthly clipboard ritual, operators now walk their own zone weekly with a phone — the walk takes about twelve minutes. Every observation is a photo, not a tick-mark, so the score finally describes the floor instead of flattering it. Zone scores dropped 20 points in week one. That was the honest baseline the paper system had been hiding.
AI names the problem in the zone, not the meeting
When an operator photographs a coolant leak or a blocked fire exit, LeanOS AI suggests the defect type and the 5S pillar it breaks in about a second — the operator confirms or edits, and the defect lands on the board with an owner and a due date. Nothing waits for the weekly meeting anymore. In the first month the plant logged 312 defects; the paper system had averaged 40 a month, most of them repeats nobody tracked.
Every defect gets an owner before the shift ends
For the stubborn ones, the AI drafts a fix-plan from what the plant has already tried on similar defects — who should own it, what the countermeasure could be, what proof of closure looks like. The maintenance head stopped being a human router of WhatsApp complaints; his queue is now a ranked list with photos, severity, and after-photo proof required to close.
The rejection problem finally got a fishbone
The creeping OEM rejections became the plant's first A3 case. The team ran the 6M fishbone with AI Diagnose, drilled a 5-Why thread into the measurement branch, and found the real cause: a batch of thread gauges past calibration that inspection kept using because replacements took six weeks to arrive. The countermeasure set — calibration register, gauge quarantine shelf, reorder trigger — closed the case in three weeks. Rejections from that customer fell 44% over the next two quarters.
The suggestion box moved into everyone's pocket
Operators now raise improvement ideas from the same phone they audit with — many in Marathi or Hindi, which LeanOS translates to English for the review board automatically. Each approved kaizen carries an expected saving; each verified kaizen carries a finance-checked actual. Twelve months in, the ledger shows ₹42 lakh a year in verified savings — a number the CFO signs, not a number the lean team hopes.
Fixes became standards, standards became skills
Every closed A3 now ends in a one-point lesson. The gauge-calibration OPL was acknowledged by all 18 inspectors inside a week, and each acknowledgement updated their row in the skill matrix. When the customer's auditor asked how the plant prevents recurrence, the quality head opened the OPL, the ack list, and the skill matrix on one screen.
The night shift started competing
Audit streaks, closed defects, and verified kaizens all earn points. The B-shift crew — historically the quiet one — topped the leaderboard in month three, and the plant head now opens the Monday review with the podium instead of the complaint list. Ideas per operator tripled. The loop keeps turning because people can see their own score turning it.
The results — safety, quality, delivery, cost, morale
Weekly walks catch blocked exits, missing guards, and oil on the floor before they become incidents — two LTIs the year before LeanOS, zero since.
The calibration A3 plus photo-verified fixes cut the key OEM's rejections nearly in half over two quarters.
Fewer rework loops and fewer line stops from repeat breakdowns pushed OTIF from 89% to 98%.
Finance-verified kaizen ledger — coolant recovery, setup-time cuts, and scrap reduction are the top three lines.
From 0.4 to 1.3 ideas per operator per quarter, with the leaderboard podium opening every Monday review.
“For a year our audits found the same oil leak and nothing happened. Now a defect photographed at 9 am has an owner by 9:05, and my CFO signs the savings number.”
Illustrative deployment story: a composite of typical LeanOS rollouts; company and names are fictional.